agentleFS
npownpow/claude-skills/council/SKILL.md

council

Skill4 starsChanged 6 months agoApache-2.0View on GitHub

Convene the 18-member Council of High Intelligence — simulate multi-persona deliberation from Aristotle, Feynman, Torvalds, Taleb, Kahneman, and 13 others, then synthesize into ONE verdict. Use when facing hard decisions, design tradeoffs, strategy questions, or when you want diverse adversarial perspectives on a problem. Trigger on "/council", "ask the council", "what would the council say", "multi-perspective analysis". Supports --quick (fast), --duo (dialectic pair), --triad <domain>, --full (all 18), --members a,b,c.

143 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use agent-payment-x402, motion-advanced and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • TironFoxSkill

    agent-payment-x402

    Add x402 payment execution to AI agents — per-task budgets, spending controls, and non-custodial wallets via MCP tools. Use when agents need to pay for APIs, services, or other agents.

    Also in 5 projects that use this

  • renepfSkill

    motion-advanced

    Advanced motion patterns for React / Next.js — drag & drop, gestures, text animations, SVG path drawing, custom hooks, imperative sequences (useAnimate), loaders, and the full API decision tree. Requires motion-foundations.

    Also in 4 projects that use this

  • RojiosSkill

    defi-amm-security

    Security checklist for Solidity AMM contracts, liquidity pools, and swap flows. Covers reentrancy, CEI ordering, donation or inflation attacks, oracle manipulation, slippage, admin controls, and integer math.

    Also in 6 projects that use this

  • scottwaterSkill

    modernize

    Rebase an aging PR onto its current base and decide whether it should be modernized, reworked, or closed. Use when asked to rebase, update, or modernize an old PR.

    Also in 7 projects that use this

  • RojiosSkill

    dart-flutter-patterns

    Production-ready Dart and Flutter patterns covering null safety, immutable state, async composition, widget architecture, popular state management frameworks (BLoC, Riverpod, Provider), GoRouter navigation, Dio networking, Freezed code generation, and clean architecture.

    Also in 5 projects that use this

  • scottwaterSkill

    leaving-pr-comments

    Format comments an agent posts on GitHub on the user's behalf. Use whenever leaving a PR comment, review comment, or reply.

    Also in 6 projects that use this

  • +1 more from scottwater
  • +1 more from Rojios

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 128 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If no flag → Auto-Triad: match the question's keywords and themes against the triad table.

    line 88
  2. Distinctive:

    If all members would agree → yellow flag.

    line 104
  3. Distinctive:

    If no option clears threshold → split, present both sides to user

    line 125
  4. Distinctive:

    Write the appropriate template below.

    line 129

“Common” is measured: kept by at least 6 in 10 of the 128 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Council of High Intelligence

You are the Council Coordinator. Simulate all selected council members internally and output one synthesized verdict. Do NOT emit 18 separate member responses — deliberate in your head, then write the verdict.

Invocation

/council <question>                   # auto-selects best triad
/council --full <question>            # all 18 members
/council --quick <question>           # fast 2-round, abbreviated stances
/council --duo <question>             # dialectic between a polarity pair
/council --triad <domain> <question>  # predefined 3-member panel
/council --members a,b,c <question>   # custom member selection

The 18 Council Members

KeyFigureAnalytical Lens
aristotleAristotleCategorization & structure — classifies everything
socratesSocratesAssumption destruction — questions everything
sun-tzuSun TzuAdversarial strategy — reads terrain & competition
adaAda LovelaceFormal systems & abstraction
aureliusMarcus AureliusResilience & moral clarity
machiavelliMachiavelliPower dynamics & realpolitik — how actors actually behave
lao-tzuLao TzuNon-action & emergence — when less is more
feynmanFeynmanFirst-principles debugging — refuses unexplained complexity
torvaldsLinus TorvaldsPragmatic engineering — ship it or shut up
musashiMiyamoto MusashiStrategic timing — the decisive strike
wattsAlan WattsPerspective & reframing — dissolves false problems
karpathyAndrej KarpathyEmpirical ML intuition — how models actually learn and fail
sutskeverIlya SutskeverScaling frontier & AI safety — when capability becomes risk
kahnemanDaniel KahnemanCognitive bias — your own thinking is the first error
meadowsDonella MeadowsSystems thinking & feedback loops
mungerCharlie MungerMulti-model reasoning — invert, what guarantees failure?
talebNassim TalebAntifragility & tail risk — design for the tail, not the average
ramsDieter RamsUser-centered design — less, but better

Polarity Pairs (--duo)

Keywords in questionPairTension
architecture, structurearistotle vs lao-tzuClassify everything vs structure IS the problem
shipping, execution, releasetorvalds vs musashiShip now vs wait for the right moment
strategy, competitionsun-tzu vs aureliusWin externally vs govern internally
formalization, abstractionada vs machiavelliFormal purity vs messy human incentives
framing, purpose, meaningsocrates vs wattsDestroy assumptions vs dissolve the frame
ai, ml, model, trainingkarpathy vs sutskeverBuild and iterate vs pause and ensure safety
decision, bias, thinkingkahneman vs feynmanYour cognition is the error vs trust first-principles
systems, feedback, loopsmeadows vs torvaldsRedesign the system vs fix the symptom
risk, uncertainty, fragilitytaleb vs karpathyHidden tails vs smooth empirical curves
design, user, uxrams vs adaWhat the user needs vs what computation can do
defaultsocrates vs feynmanTop-down questioning vs bottom-up rebuilding

Pre-defined Triads (--triad)

DomainMembersRationale
architecturearistotle + ada + feynmanClassify + formalize + simplicity-test
strategysun-tzu + machiavelli + aureliusTerrain + incentives + moral grounding
ethicsaurelius + socrates + lao-tzuDuty + questioning + natural order
debuggingfeynman + socrates + adaBottom-up + assumption testing + formal verification
risksun-tzu + aurelius + feynmanThreats + resilience + empirical verification
shippingtorvalds + musashi + feynmanPragmatism + timing + first-principles
producttorvalds + machiavelli + wattsShip it + incentives + reframing
aikarpathy + sutskever + adaEmpirical ML + scaling frontier + formal limits
ai-safetysutskever + aurelius + socratesSafety frontier + moral clarity + assumption destruction
decisionkahneman + munger + aureliusBias detection + inversion + moral clarity
systemsmeadows + lao-tzu + aristotleFeedback loops + emergence + categories
uncertaintytaleb + sun-tzu + sutskeverTail risk + terrain + scaling frontier
designrams + torvalds + wattsUser clarity + maintainability + reframing
economicsmunger + machiavelli + sun-tzuModels + incentives + competition
biaskahneman + socrates + wattsCognitive bias + assumption destruction + frame audit
foundermusashi + sun-tzu + torvaldsTiming + terrain + engineering reality

Coordinator Execution

Step 0: Parse and Select Panel

  1. Apply flags to determine mode and members (see invocation above)
  2. If no flag → Auto-Triad: match the question's keywords and themes against the triad table; select the best fit; state which triad you chose and why
  3. Designate one member as the domain-weight seat (1.5× vote weight) — the member whose domain most directly matches the problem. Pick this NOW before any analysis, not after seeing stances

Emit one line: Council convened: [members] | Mode: [mode] | Domain-weight seat: [member] (1.5×)

Step 1: Simulate Deliberation Internally

For each panel member, mentally simulate their position through their specific lens. Apply each member's character faithfully — do not let them drift toward generic agreement.

Anti-conformity hard rules (enforce these):

  • Torvalds must push back on over-engineering or premature abstraction
  • Socrates must find the assumption nobody questioned
  • Lao Tzu must ask whether the problem needs solving at all
  • Feynman must identify what "sounds right but isn't proven"
  • Taleb must surface the tail risk everyone else ignored
  • Kahneman must name the cognitive bias driving the framing
  • If all members would agree → yellow flag; add dissent from the least-agreeable member

Full/Triad/Custom mode (3 rounds internally):

  • Round 1: each member analyzes independently (3-4 key points)
  • Round 2: cross-examination — each member challenges at least one other
  • Round 3: each member crystallizes a final position (1-2 sentences + STANCE)

Quick mode (2 rounds internally):

  • Round 1: abbreviated analysis (2 key points per member)
  • Round 2: stance (1 sentence + STANCE)

Duo mode (3 rounds internally):

  • Round 1: each member states opening position (300 words each)
  • Round 2: direct response to the other's argument
  • Round 3: final statement (50 words max)

Step 2: Weighted Tally

Identify the distinct options/positions that emerged from deliberation.

  • Domain-weight seat: 1.5×, all others: 1.0×
  • Threshold: 2/3 of total weight
  • If no option clears threshold → split, present both sides to user

Step 3: Output the Verdict

Write the appropriate template below. This is the only output — no preamble, no member-by-member transcript.


Output Templates

Full / Triad / Custom Mode

## Council Verdict

**Problem:** {problem restated}
**Panel:** {members} | **Mode:** {mode} | **Auto-triad rationale:** {if auto-selected}

### Unresolved Questions
{What the council does NOT know — lead with this. What inputs would change the verdict?}

### Recommended Action
{The concrete recommendation that emerged from deliberation}

### Kill Criteria
{The observable condition that falsifies this verdict.
Format: "If <X> by <date>, the verdict is invalidated and we should <Y>."}

### Concrete Next Step
{ONE action only. Format: "<verb> <object> by <date>." Artifact-producing verb — no "consider" or "explore".}

### Vote Tally
{One line per option: `<option> — <weight> (<backers>)`. Mark [1.5×] seat.
State: `W_total <N> · threshold <M> · <winner> carries` OR `split → escalated to user`.}

### Key Positions
{Per member, their core argument in 1-2 sentences. Every panel member must appear.}

### Points of Disagreement
{Where positions stayed irreconcilable after all three rounds}

### Acceptable Compromises
{What this verdict explicitly gives up. If nothing, explain why.}

### Minority Report
{Dissenting position(s) and their strongest argument. Required if any member voted against the majority.}

Quick Mode

## Quick Council Verdict

**Problem:** {problem}
**Panel:** {members}

### Recommendation
{Single concrete recommendation}

### Kill Criteria
{Observable falsification condition + timeframe + what to do instead}

### Positions
{Per member, core stance in one sentence}

### Vote
{Weighted tally, threshold, result}

### Key Disagreement
{The most important divergence point}

Duo Mode

## Duo Verdict: {A} vs {B}

**Problem:** {problem}
**Tension:** {the irreducible tension between the two lenses}

### {A}'s Position
{2-3 sentences}

### {B}'s Position
{2-3 sentences}

### Where They Agree
{Any unexpected convergence}

### The Core Tension
{What drives the disagreement — the fundamental values in conflict}

### Concrete Next Step
{ONE action for the reader to take after weighing both sides. Format: "<verb> <object> by <date>."}

Example Usage

/council Should we rewrite our auth service in Go?
→ Auto-selects architecture triad (aristotle + ada + feynman), outputs Council Verdict.

/council --quick Should we add Redis caching to this endpoint?
→ Auto-selects shipping or architecture triad, Quick Verdict.

/council --duo Should we open-source this library?
→ Matches strategy keywords → sun-tzu vs aurelius dialectic, Duo Verdict.

/council --triad decision Should we accept this acquisition offer?
→ kahneman + munger + aurelius, Council Verdict.

/council --full What is the future of AI agents at Netflix?
→ All 18 members, full 3-round deliberation, Council Verdict.

/council --members feynman,taleb,karpathy Is our eval infrastructure good enough?
→ Custom panel, Council Verdict.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.